Method for determining a standardized speed of a soft mobility vehicle

Abstract

The present invention relates to a method for determining a standardized speed of a soft mobility vehicle on at least one portion of a transport network, said transport network comprising a plurality of interconnected strands, with the following steps: a) Geolocated instantaneous speeds of a plurality of soft mobility trips of a plurality of users within said portion of the transport network are acquired; b) For each strand of said portion of the transport network, an intermediate speed model is constructed for each user having traveled said strand by means of said acquired geolocated instantaneous speeds, said intermediate speed model determining an intermediate speed for each strand for each user;(c) For each strand of said portion of transport network, a normalized speed model is constructed as the average of intermediate speeds, the normalized speed model determining a normalized speed for each strand for all users; and (d) said normalized speed is determined for each strand of said portion of transport network by minimizing a cost function which depends on a difference between an intermediate speed determined by said intermediate speed model and a normalized speed, and by applying said normalized speed model. Figure 1 to be published;

Description

Title of the invention: Method for determining a standardized speed of a soft mobility vehicle Technical field

[0001] The present invention relates to the field of determining the speed of movement of a soft mobility vehicle, in particular for travel by bicycle, scooter, on foot, etc.

[0002] In order to understand, predict and / or improve soft mobility travel within a geographical area, it is necessary to know the travel speed for these modes of transport within the geographical area. For example, the travel speed can be used to determine a route for an upcoming trip between an origin and a destination. This determined route can be used to guide the user of the transport mode. Prior art

[0003] Massive mobility data collection and analysis has taken a leap forward with the development of highly accessible wearable sensors, now included in devices such as mobile phones and smartwatches. Mobility data can be retrieved by different types of sensors, including: cellular network, Wi-Fi, Bluetooth and GPS (Global Positioning System) sensor, which was first made available on a mobile phone in the late 1990s.

[0004] Furthermore, in recent years, activity tracking devices and applications, particularly for sports applications (for example STRAVA™), have been developed and allow on the one hand training and comparison of activities, and on the other hand determining various data linked to the movements of users, in particular their speed.

[0005] Data from these different sources can be used to determine a travel speed of a mode of transport for each strand of the geographic area.

[0006] Generally an average speed and / or a median speed are calculated from the measured data. However, this average or median speed is dependent on the users. Indeed, an average speed, or a median speed, is generally calculated directly from the speeds observed on a section, they are then likely to carry the biases of the population crossing it. In fact, the average or median speed is not directly representative of the difficulty of the road but rather of the population that crosses it. For example, if the measured bicycle travel speed data comes from a sports tracking device or application, the average or median speed may be high (due to the users and the type of equipment), in particular higher than the average or median speed of bicycle users for walking or commuting. Thus, it seems difficult to compare sections with each other, if they are not traveled by the same population. If we decide to only use the roads where the average speed is the highest, we will then be led to follow the journeys of experienced users (for example cyclists), which will not necessarily result in a faster journey, if we are not an experienced user.

[0007] It is therefore necessary to determine a speed that is independent of the population traveling on a route or a section of a route. This population-independent speed makes it possible in particular to determine the fastest route between an origin and a destination for any user.

[0008] Patent application US2012 / 0053896 relates to a method for comparing performance between several users. This method compares in particular the instantaneous speed of users on the same section. However, this method does not make it possible to determine a speed independent of the population on a section of a transport network. Summary of the invention

[0009] The aim of the present invention is to determine, in a robust and reliable manner, for a portion of a transport network, a speed independent of the population crossing said portion. For this purpose, the invention relates to a method for determining a standardized speed of a soft mobility vehicle within a transport network, this method implementing a step of acquiring geolocated instantaneous speeds, an intermediate speed model, a standardized speed model and a minimization of a cost function. The intermediate speed and standardized speed models make it possible to dissociate the user's performance from a speed independent of the user. The minimization of the cost function ensures the robust and reliable determination of the standardized speed, which is independent of the population crossing the portion of the transport network.

[0010] The invention further relates to a method for determining a route which implements such a method for determining a standardized speed.

[0011] The subject of the invention is a method for determining a standardized speed of a soft mobility vehicle, in particular a bicycle, on at least one portion of a transport network, said transport network comprising a plurality of interconnected strands, such that the following steps are implemented:

[0012] a) Geolocated instantaneous speeds of a plurality of soft mobility trips of a plurality of users within said portion of the transport network are acquired;

[0013] b) For each strand of said portion of transport network, an intermediate speed model is constructed for each user having traveled said strand by means of said acquired geolocated instantaneous speeds, said intermediate speed model determining an intermediate speed for each strand for each user;

[0014] c) For each strand of said portion of transport network, a normalized speed model is constructed as the average of intermediate speeds, the normalized speed model determining a normalized speed for each strand for all users; and

[0015] d) Said normalized speed is determined for each strand of said portion of transport network by minimizing a cost function which depends on a difference between an intermediate speed determined by said intermediate speed model and a normalized speed, and by applying said normalized speed model.

[0016] The invention also relates to a method for determining a route of a soft mobility vehicle, in particular a bicycle, between an origin and a destination belonging to a portion of a transport network, such that the following steps are implemented:

[0017] v) A normalized speed is determined for each strand of said portion of transport network by means of the normalized speed determination method described above; and

[0018] w) A path between said origin and said destination is determined by means of a shortest path algorithm taking into account said determined normalized speed.

[0019] The characteristics and advantages of the method according to the invention will appear on reading the following description of non-limiting examples of embodiments, with reference to the appended figures described below. List of figures

[0020] [Fig.l]

[0021] [Fig.l] illustrates the steps of the method according to a first embodiment of the invention.

[0022] [Fig.2]

[0023] [Fig.2] illustrates the steps of the method according to a second embodiment of the invention.

[0024] [Fig.3]

[0025] [Fig.3] illustrates the steps of the method according to a third embodiment of the invention.

[0026] [Fig.4]

[0027] [Fig.4] illustrates a portion of a transport network with three movements for an example application of the invention.

[0028] [Fig.5]

[0029] [Fig.5] illustrates a graph of the coefficient of performance of several movements of several users.

[0030] [Fig.6]

[0031] [Fig.6] illustrates on the graph of [Fig.5] with the performance coefficient of each user for all the journeys. Description of the embodiments

[0032] The invention relates to a method for determining a standardized speed of a soft mobility vehicle within a transport network. A soft mobility vehicle is a vehicle that uses human energy and, consequently, contributes to a reduction in CO2 emissions, particularly compared to thermal vehicles. This may include walking, running, cycling, scootering, roller skating, or any similar type of travel. A transport network is a set of roads and paths in a predefined geographical area. This predefined geographical area may be a district of a city, a town, a community of municipalities, a department, etc. A strand of the transport network is an elementary subdivision of the transport network between two consecutive nodes of the transport network.For example, a strand of the transport network may be a road between two consecutive intersections, between two consecutive signals, between an intersection and a signal, or a section of a motorway between two consecutive exits, etc. Thus, we have a fine division of the transport network, and a model that is adapted to the transport network without microscopic data. For example, for the implementation of the invention for bicycle travel, the transport network may include roads, cycle paths, paths, etc., but not motorways.

[0033] A normalized speed is a speed independent of the user of the soft mobility vehicle. This normalized speed can be defined as the speed of travel in a soft mobility vehicle for an average user (i.e. a user of standard, average performance / physical capacity). The normalized speed is representative of the difficulty of the strand of the transport network, independently of the physical capacities of the user. At the scale of a strand, the normalized speed can be defined by the following formula: X[ = ttjVjj with aJ a performance coefficient of user j, vij an instantaneous speed of strand i by user j, Xg a normalized speed of strand i. This normalized speed can be obtained by the process steps described in the remainder of the description.

[0034] The transport network can be represented by a graph, called a transport graph. The transport graph is composed of a set of edges and nodes, where the nodes can represent intersections, and the edges can represent the portions of roads / paths (strands) between the intersections. The transport graph can be obtained from an online mapping service ("webservice"), for example Here™ (Here Apps LLC, Netherlands) which provides the edges of the graph as pure geometric objects. Preferably, the transport graph is consistent with the transport network (all physical connections between two roads, and only these, are represented by the nodes of the graph), and as time-invariant as possible. In addition, the transport graph can be simplified, by not taking into account portions of roads such as dead ends, motorways or expressways, depending on the type of vehicle considered.

[0035] The present invention therefore relates to a method for determining a standardized speed of a soft mobility vehicle within a transport network. This method implements the following steps:

[0036] 1. Acquisition of geolocated instantaneous speeds

[0037] 3. Construction of an intermediate speed model

[0038] 4. Construction of a normalized speed model

[0039] 5. Determination of the normalized speed

[0040] These steps can be implemented by computer means, for example a computer or a server. These steps will be detailed in the remainder of the description.

[0041] [Fig.l] illustrates, schematically and in a non-limiting manner, the steps of the method according to a first embodiment of the invention. Firstly, geolocated instantaneous speeds are acquired (ACQ). Then, an intermediate speed model (MVI) and a normalized speed model (MVN) are constructed. These two models and the geolocated instantaneous speeds are used in a minimization method (MIN) to determine the normalized speed (X).

[0042] According to one embodiment, the method may further comprise a step of preprocessing the acquired geolocated instantaneous speeds. Thus, it is possible to limit the noise in the acquired speeds, and to best respect speed distribution assumptions. Advantageously, this preprocessing step may comprise the filtering of the acquired speeds. This filtering may concern the filtering of speeds that are too slow and / or too fast according to a non-limiting example, the filtering of speeds below 10 km / h and speeds above 40 km / h. Thanks to this filtering, it is possible to retain within the acquired speeds only those related to a moving vehicle, and it is possible to avoid including speeds acquired from a trip made in a motorized mode of transport.

[0043] For this embodiment, the method may comprise the following steps:

[0044] 1. Acquisition of geolocated instantaneous speeds

[0045] 2. Filtering of geolocated instantaneous speeds

[0046] 3. Construction of an intermediate speed model

[0047] 4. Construction of a normalized speed model

[0048] 5. Determination of the normalized speed

[0049] These steps can be implemented by computer means, for example a computer or a server. These steps will be detailed in the remainder of the description.

[0050] [Fig. 2] illustrates, schematically and in a non-limiting manner, the steps of the method according to a first embodiment of the invention. The steps identical to the embodiment of [Fig. 1] are not detailed again. The method further comprises a step of filtering (FIL) the geolocated instantaneous speeds (ACQ). The step of determining the normalized speed by minimization (MIN) takes into account the filtered geolocated instantaneous speeds.

[0051] Acquisition of geolocated instantaneous speeds

[0052] During this step, geolocated instantaneous speeds of a plurality of soft mobility trips of a plurality of users within the portion of the transport network are acquired. Thus, the method according to the invention requires several speeds of several trips and speeds of several users. The instantaneous speed is the speed of the vehicle at a point in the transport network; it is a point speed.

[0053] According to an implementation of the invention, the geolocated instantaneous speeds can be acquired by means of measurements by a geolocation sensor. A geolocation sensor is a sensor measuring the position and speed of the user, as well as possibly the time of the measurement. This is then referred to as geolocated speed. Such a geolocation sensor can use GPS technology (from the English "Global Positioning System" which can be translated as global positioning system), Galileo technology or any equivalent technology. Advantageously, the geolocation sensor can be provided in a smartphone of the user, or in a connected watch, or any connected object or any similar object.For this implementation of the invention, the method can therefore comprise a prior step of measuring the position and speed of several movements of several users by means of a geolocation sensor, in particular a geolocation sensor per user.

[0054] Advantageously, for this step, thanks to geolocation, we can associate each instantaneous speed with a strand of the transport network. Thus, we can have for several strands of the transport network one or more instantaneous speeds of one or more journeys of one or more users. For this association, we can put in implements a "mapmatching" method (which can be translated as "cartospondence" or geographic correspondence). For this cartospondence algorithm, we can use a geographic information system GIS (or in English GIS for "Geography Information System"). Here Maps™, Google Maps™, OpenStreetMap™, etc. are examples of a GIS geographic information system. Such a GIS geographic information system may contain the dedicated lanes of soft mobility modes of transport, for example cycle paths, paths, etc. If several instantaneous speeds of the same user's movement are associated with a single strand, then an average speed for this strand and this user can be determined for this strand. According to one aspect of the invention, the passages of users can be separated, that is to say that if the user passes twice at the same place on two different occasions (or even if he makes a loop), several average speeds are obtained. On the contrary, this is also a problem in the development of the overall average speed calculated by the methods of the prior art, because it is biased by the users “chronic” who always travel the same route every day for example, and who therefore count for a very significant weight in the calculation.

[0055] 2, Filtering of geolocated instantaneous speeds

[0056] During this optional step, the geolocated instantaneous speeds acquired during step 1 are filtered. Thus, it is possible to limit the noise in the acquired speeds, and to best respect speed distribution assumptions. This filtering may concern the filtering of speeds that are too slow or too fast; according to a non-limiting example, the filtering of speeds below 10 km / h and speeds above 40 km / h may be carried out for a bicycle travel mode. Thanks to this filtering, only those speeds related to a moving vehicle can be retained within the acquired speeds, and it is possible to avoid including speeds acquired from a trip made in a motorized transport mode. This step may include other similar preprocessing operations, for example the elimination of outliers.

[0057] Advantageously, the possible association of instantaneous speeds with the strands of the road network can be implemented for the filtered instantaneous speeds.

[0058] In the same way, the possible determination of an average speed of a section of the road network for a user can be implemented for the filtered instantaneous speeds.

[0059] 3. Construction of an intermediate speed model

[0060] During this step, for each strand of the transport network portion, an intermediate speed model is constructed for each user having traveled the strand considered. The intermediate speed model determines an intermediate speed for each transport network strand for each user having traveled the strand considered. An intermediate speed is a speed representative of a user on a section of the transport network. In other words, the intermediate speed can be defined as the correction of the measured speed of a user with the performance factor, corresponding to a performance gap of this user on this particular section. In other words, the intermediate speed corresponds to the local performance speed of the user, which corresponds to the performance gap between this section of road and the rest of his or her journeys.

[0061] According to one embodiment of the invention, said intermediate speed model can be constructed using the following formula: Xÿ — CLjVjj with Xy an intermediate speed of strand i for user j, aJ a performance coefficient of user j, an instantaneous speed of strand i by user j acquired or even filtered in the previous steps. 4. Construction of a normalized speed model

[0062] In this step, for each strand of the transport network portion, a normalized speed model is constructed as the average of the intermediate speeds of the different users. The normalized speed model determines a normalized speed for each strand for all users.

[0063] According to one embodiment, the normalized speed model can be constructed using the following formula: Xj = ^IbjWiiXij with %i a normalized speed of strand i, Xy an intermediate speed of strand i for user j, ni a number of users having traveled strand i, wij a coefficient equal to 1 if user j has traveled strand i and equal to 0 otherwise. 5. Determination of the standardized speed

[0064] During this step, the normalized speed is determined in at least one strand of the transport network by minimizing a cost function representing, for a plurality of strands of the transport network, a difference between an intermediate speed obtained by the intermediate speed model constructed in step 3, and a normalized speed. A cost function designates a function which serves as a criterion for determining the best solution to an optimization problem. In this case, it is a question of optimizing the normalized speed on a plurality of strands of the transport network. After minimization, the normalized speed model determined in step 4 is applied to deduce the normalized speed.

[0065] According to one embodiment, the cost function L to be minimized can be written as follows:

[0066] L = - Xj ) with %i a normalized speed of strand i, Xy an inter speed median of strand i for user j, and f denotes a function, for example a weighting.

[0067] According to one implementation, the cost function L to be minimized can be written as follows:

[0068] L = w- (av - X)2 with i a normalized speed of strand i, Xy an intermediate speed of strand i for user j, wij a coefficient equal to 1 if user j has traveled strand i and equal to 0 otherwise.

[0069] with aj a performance coefficient of user j, vij an instantaneous speed of strand i by user j, X, a normalized speed of strand i.

[0070] Minimizing this cost function makes it possible to determine the performance coefficients aJ for each user. Then, the normalized speed of each strand of the portion of the transport network can be determined using the normalized speed model and the acquired geolocated instantaneous speeds: X = 4-IL w,.-»py ,• 7 J DD J

[0071] The minimization of this cost function can be implemented by any method, for example by a simplex method, BFGS (Broyden-Fletcher-Goldfarb-Shanno method), or conjugate gradient.

[0072] In the remainder of the description, a non-limiting example of minimization of the cost function is described:

[0073] Using the variance formulas we can write:

[0074] L = £.[£. (WlJ(ajViX)]

[0075]

[0076] We can ask

[0077] = diag ( )

[0078] B =

[0079] C = diag(^)

[0080] D = 2(A-BtCB)

[0082] With n the number of users

[0083] We can then write g — iaT[)a

[0084] The gradient of this formula can be written as L — Da .

[0085] The equality constraint can be written on each connected graph g in the manner next: with Yij = Xij

[0086] The gradient of Kg can be written:

[0087] A condition for a to be a minimal of L while verifying the G equality constraints can be written: 7 € r7 VL - F LV.iï5

[0088] This condition can also be written:

[0089] These equations are independent and can be solved separately.

[0090] If the matrix Dg is invertible, the unique solution of the g-th equation can be written:

[0091] With which can be determined by: ■ i p.; . ' ' ■ »

[0092] If the matrix Dg is non-invertible, the solution to the problem can be {xeK® U a < = I4Ç), if the dimension of Ker(Dg) is 1 this solution is unique, if this dimension is greater than 1, then we can arbitrarily choose a direction of Ker(Dg).

[0093] According to an embodiment, in which a performance coefficient of each user is determined, it is also possible to deduce the average predictive speed of a user on a strand of the transport network portion by means of the user's performance coefficient and the normalized speed of the strand considered. The average predictive speed of a user is the predicted speed of movement of the user on the strand. Thus, the predictive speed of the user can be determined even on a strand of the transport network portion that the user has not traveled, this speed being representative of the user's capacity and equipment. In this way, a travel time can be predicted for the user for all the strands of the road transport portion.

[0094] Advantageously, said average predictive speed of at least one user on a strand can be deduced using the following formula:,. _ A with Vmoyjj the ~~ aj average predictive speed of user j on strand i, the normalized speed of strand i and a> the performance coefficient of user j.

[0095] Preferably, this implementation can be applied for all strands of a transport network.

[0096] According to an embodiment option, the normalized speed can be determined and / or the average predictive speed of a user on a transport network different from the one for which the geolocated instantaneous speed measurements are acquired. Thus, it is possible to determine a speed (normalized or average predictive of a user) of a soft mobility trip even for a transport network for which it is difficult to acquire geolocated instantaneous speeds. In this way, it becomes possible to determine a travel time even for a transport network with few speed measurements. For this implementation option, the following steps can be implemented: A. Building a learning base B. Construction of a model of a user's average instantaneous or predictive speed C. Determining the instantaneous or average speed of a user

[0097] These steps can be implemented by computer means, in particular a computer or server. Steps 1 and 2 can be performed offline, and step 3 can be performed online. Using the offline-built model online allows for reduced online computation time and limited computer memory requirements compared to a microscopic model, while still achieving good accuracy. These steps are detailed in the following description.

[0098] For step A, any variant embodiment of the invention can be applied to a learning transport network (i.e. to the transport network on which geolocated instantaneous speeds are acquired), and thus obtain an instantaneous learning speed and possibly an average predictive speed of a learning user on all the strands of the learning transport network.

[0099] The learning base is constructed from at least macroscopic data of a learning road network and from the normalized and / or average predictive speeds of a user. The macroscopic data of the road network correspond to information related to the road network, such as infrastructure, slope, signage, etc. The traffic data is the measured or simulated data representing the traffic on the road network.

[0100] According to one aspect of the invention, the macroscopic data of the transport network may be topology (i.e. slope, curvature of the road, intersections, signage, etc.), type of road strand (e.g. urban road, protected or unprotected cycle lane, path, etc.), number of lanes of the road strand, maximum speed of the road strand, signage of the road strand, slope of the road strand, length of the road strand, typology of vehicles accepted, infrastructure present on the strand that could influence driving (bus stops, gas stations, signage, etc.), etc. Preferably, the macroscopic data of the transport network can be provided by a Geographic Information System (GIS). Here Maps™, Google Maps™, OpenStreetMap™ are examples of Geographic Information Systems. Macroscopic data are always available and from any location. Thus, they can serve as unique inputs to the training base and the pollutant emissions model.

[0101] For step B, a speed model (normalized or average predictive of a user) is constructed by a machine learning method trained using the learning base constructed as described above, comprising at least macroscopic data from the learning road network and the speed determined on the road network. The speed model associates with at least one strand of the road network, a speed (normalized or average predictive of a user) as a function of the macroscopic data. For example, a normalized speed model can be written:

[0102] X = g(Typ, Slope, long) with X the normalized speed, Typ the type of the strand, Slope the slope of the strand, long the length of the strand.

[0103] The function g may be based on a supervised learning algorithm, such as multivariate linear regression, a neural network, a random forest, any analogous method, or a combination of these methods. As a non-limiting example, the method may implement a neural network with two hidden layers of twenty and fifteen neurons respectively, or may implement a random forest with twenty estimators (or trees) of depth equal to 15, or any equivalent algorithm.

[0104] According to one aspect, this step may include a validation method, preferably a cross-validation method, in particular a k-fold cross-validation method. This cross-validation method makes it possible to reduce “overfitting” problems and to improve the accuracy of the model.

[0105] During step C, the speed (normalized or average predictive of a user) is determined on at least one strand of the transport network considered (different from the learning road network), by applying the speed model constructed in step B to macroscopic data of the transport network strand considered. In other words, for the transport network considered, the macroscopic data are known, and the model constructed in the previous step is applied to these data relating to the strand of the transport network considered, which makes it possible to determine a speed (normalized or average predictive of a user) for a soft mobility trip for this strand of the transport network.

[0106] Furthermore, the invention relates to a method for determining a route of a soft mobility vehicle, in particular a bicycle, between an origin and a destination belonging to a portion of a transport network. For this method, it is possible to implement implement the following steps:

[0107] v) A normalized speed is determined for each strand of said portion of transport network by means of the normalized speed determination method according to any one of the variants or combinations of variants described above; and

[0108] w) A path between the origin and the destination is determined by means of a shortest path algorithm taking into account said normalized speed determined in the previous step.

[0109] This method makes it possible to identify an optimal route for a soft mobility vehicle, by taking into account a speed that is truly representative of each strand of the transport network. The determination of the route can be implemented in particular for the purposes of guiding (“routing”) a user, in particular by means of a smartphone, a connected object, etc. Indeed, in this case, the user can use the smartphone or the connected object to acquire the origin and / or the destination, and / or to carry out the guidance.

[0110] Thus, for this embodiment, the method can implement the following steps:

[0111] 1. Acquisition of geolocated instantaneous speeds

[0112] 2. Filtering of geolocated instantaneous speeds (optional)

[0113] 3. Construction of an intermediate speed model

[0114] 4. Construction of a normalized speed model

[0115] 5. Determination of the normalized speed

[0116] 6. Determination of a route.

[0117] These steps can be implemented by computer means, for example a computer or a server.

[0118] [Fig. 3] illustrates, schematically and in a non-limiting manner, the steps of the method of this embodiment. The steps identical to the embodiment of [Fig. 2] are not detailed again. The filtering step remains optional for this embodiment. The method further comprises a step of acquiring an origin and a destination (O / D) of a route, as well as a step of determining a route (TRA) as a function of the origin and destination (O / D) and the normalized speed (X).

[0119] The shortest path algorithm minimizes a criterion, in particular the travel time, by taking into account the normalized speed of each strand of the transport network portion (the normalized speed can be seen as a weighting of each strand of the transport network portion). According to an embodiment option, the shortest path algorithm used can be chosen from the Bellman-Ford algorithm or the Djikstra algorithm or any similar algorithm.

[0120] According to an embodiment option, the shortest path algorithm can further take into account the user's performance coefficient. Thus, the optimal path determined is dependent on the user's ability.

[0121] In addition, the method according to the invention may include a step of displaying the standardized speed. During this step, the standardized speed determined is displayed on a transport map, in particular a road map. This display may take the form of a note or a color code or a thickness of representation of the road. This display may be carried out on board the vehicle: on the dashboard, on a portable autonomous device, such as a geolocation device (GPS type), a mobile phone (smartphone type). It is also possible to display the standardized speed on a website. In addition, the standardized speed may be shared with public authorities (for example, road manager) and public works companies.In this way, public authorities and public works companies can determine which roads have a low or high standard speed and adapt the roads to users (e.g. creation of new lanes, modification of signage, etc.).

[0122] The invention also relates to a method for managing infrastructure of a transport network. For this method, the following steps can be implemented:

[0123] x) At least one normalized speed is determined by means of a normalized speed determination method according to any one of the variants or combinations of variants described above; and

[0124] y) At least one infrastructure of the transport network is modified according to the determined standardized speed, for example an infrastructure for which the standardized speed is maximum, minimum, lower or higher than a predetermined threshold.

[0125] Thus, a transport network can be managed to limit or even avoid traffic jams, accident risks, etc.

[0126] According to one embodiment, the modification of the infrastructure can be chosen in particular from the addition of signage (speed limit, traffic light, give way, stop, etc.), the construction of a new lane (for example a cycle lane), passage of a one-way strand, construction of a new road, etc. Examples

[0127] The characteristics and advantages of the method according to the invention will appear more clearly on reading the examples below.

[0128] The first example concerns a bicycle transport network. This transport network has 6 strands. This transport network is used by three cyclists (users) of different conditions.

[0129] [Fig.4] illustrates, schematically and in a non-limiting manner, the transport network of this example. On this section, we represent: - in thick lines the journey of a first cyclist,

[0130]

[0131] - in dotted lines the route of a second cyclist, - alternating dashes and dots, the route of a third cyclist. The sections (strands) are identified by letters from a to f. The instantaneous speeds of the first cyclist are noted VI, the instantaneous speeds of the second cyclist are noted V2 and the instantaneous speeds are noted V3. Table 1 indicates for each strand the speed of each cyclist, the average speed Vmoy (the average of the speeds of the users) and the normalized speed X determined by the method according to the invention. The speeds are expressed in km / h. [Tables 1] Strand VI V2 V3 Vavg

[0132] It can be noted that the normalized speed is different from the average speed, and is independent of the level of the user who travels the strand. For example, the first cyclist is the fastest, and the third cyclist is the slowest. Consequently, the average speed of a strand traveled by only one of the two (in this case strands b and f) does not give a representative indication of a speed independent of the user, unlike the normalized speed.

[0133] In addition, for this example, the coefficient of performance of each user is determined. For the first user, the coefficient of performance is 0.72, for the second user, the coefficient of performance is 1.04 and for the third user the coefficient of performance is 1.6.

[0134] For the second example, we consider 5 cyclists making different journeys over a period of 6 months. We apply the method according to the invention to determine a performance coefficient for each user for each journey.

[0135] [Fig.5] illustrates a graph of the coefficient of performance a as a function of the average speed of the user on the route V in km / h. Each point therefore represents a route of a user. Each user is illustrated by a particular shape: plus, cross, diamond, circle and square. On this graph, we can notice that each user is found in an area of ​​the graph; there is a large variation in the coefficient of performance, but a variation which is strongly correlated with his speed on the journey (the same user generally making a more or less identical journey).

[0136] [Fig.6] is a graph similar to [Fig.5] on which we also represent with a point of larger size (having the corresponding shape of the user: plus, cross, diamond, circle and square) the coefficient of performance and the average speed of each user for all of these journeys.

[0137] If we compare users with each other (performance coefficient at the user level without separating their journeys), we can see a hierarchy between them, even at the same average speed, for example the points represented by pluses and crosses ([Fig.5]) have a significant average speed of 18km / h over their entire journey but a different performance coefficient, this is explained because the user represented by crosses actually takes segments considered easier by the algorithm (higher normalized speed). By comparing users with the performance coefficient, we can create a hierarchy that respects their speed (a faster user seems more efficient than a slower one) but also takes into account the difficulty of the paths they take.

Claims

Claims

1. Method for determining a standardized speed of a soft mobility vehicle, in particular a bicycle, on at least one portion of a transport network, said transport network comprising a plurality of interconnected strands, characterized in that the following steps are implemented: a) Geolocated instantaneous speeds of a plurality of soft mobility trips of a plurality of users within said portion of the transport network are acquired; b) For each strand of said portion of the transport network, an intermediate speed model is constructed for each user having traveled said strand by means of said acquired geolocated instantaneous speeds, said intermediate speed model determining an intermediate speed for each strand for each user;(c) For each strand of said portion of transport network, a normalized speed model is constructed as the average of intermediate speeds, the normalized speed model determining a normalized speed for each strand for all users; and (d) said normalized speed is determined for each strand of said portion of transport network by minimizing a cost function which depends on a difference between an intermediate speed determined by said intermediate speed model and a normalized speed, and by applying said normalized speed model.;

2. Method according to claim 1, wherein said intermediate speed model is constructed by means of the following formula: Xÿ = djVij with Xÿ an intermediate speed of strand i for user j, aî a performance coefficient of user j, vij an instantaneous speed of strand i by user j.

3. Method according to one of the preceding claims, in which said normalized speed model is constructed by means of the following formula: X; = with a normalized speed of strand i, Xÿ an intermediate speed of strand i for user j, nî a number of users having traveled strand i, wij a coefficient equal to 1 if user j has traveled strand i and equal to 0 otherwise.

4. Method according to one of the preceding claims, in which said cost function is written: £ — y ( a v- - X-)2 with said function cost, wij a coefficient worth 1 if user j has traveled along strand i and worth 0 otherwise, aJ a performance coefficient of user j, vij an instantaneous speed of strand i by user j, X, a normalized speed of strand i.

5. Method according to one of the preceding claims, in which a performance coefficient of at least one user is further determined, and an average speed of the at least one user on a section of said portion of the road network is deduced therefrom by means of said normalized speed.

6. Method according to claim 5, in which said average speed of the at least one user on a strand is deduced by means of the following formula: with vmoy,ij the average speed of a; user j on strand i, the normalized speed of strand i and aJ the performance coefficient of user j.

7. Method according to one of the preceding claims, in which said method further comprises a step of preprocessing the acquired geolocated instantaneous speeds, in particular by filtering the minimum and / or maximum speeds.

8. Method according to one of the preceding claims, in which said geolocated instantaneous speeds are acquired by means of measurements by a geolocation sensor, and each instantaneous speed is associated with a strand of said transport network.

9. Method for determining a route of a soft mobility vehicle, in particular a bicycle, between an origin and a destination belonging to a portion of a transport network, characterized in that the following steps are implemented: v) A standardized speed is determined for each strand of said portion of transport network by means of the standardized speed determination method according to one of the preceding claims; and w) A route between said origin and said destination is determined by means of a shortest path algorithm taking into account said determined standardized speed.

10. Method according to claim 9 implementing the method for determining normalized speed according to claim 5 or 6, wherein said shortest path algorithm takes into account said coefficient of performance of said user.

Images